Tour of Go rot13Reader buffer not being updated after Read function finishes - go

Here is my implementation of the exercise using strings.Map (the rot13 function is straight out of golang's docs). The issue is that the buffer does not seem to be modified after the Read function returns. Here is the code:
package main
import (
"io"
"os"
"strings"
"fmt"
)
type rot13Reader struct {
r io.Reader
}
func (reader *rot13Reader) Read(b []byte) (int, error) {
rot13 := func(r rune) rune {
switch {
case r >= 'A' && r <= 'Z':
return 'A' + (r-'A'+13)%26
case r >= 'a' && r <= 'z':
return 'a' + (r-'a'+13)%26
}
return r
}
n, err := reader.r.Read(b)
result := []byte(strings.Map(rot13, string(b)))
b = []byte(result)
fmt.Println(string(b))
return n, err
}
func main() {
s := strings.NewReader("Lbh penpxrq gur pbqr!")
r := rot13Reader{s}
io.Copy(os.Stdout, &r)
}
and the output:
You cracked the code!
Lbh penpxrq gur pbqr!You cracked the code!
Clearly the buffer has been modified in the Read function, but it does not seem to be the case after it returns. If I were to comment out the fmt.Println(string(b)), the output would just be:
Lbh penpxrq gur pbqr!
Is there something quirky about Readers and Writers that I should know about?

In Go, all arguments are passed by value, as if by assignment to the parameter or receiver (a shallow copy).
In Go, a slice is implemented as
type slice struct {
array unsafe.Pointer
len int
cap int
}
When the slice is passed by value, after you return, you will not see any changes you make to the copy of the struct fields. You will only see any changes to elements of the underlying array.
In your case, you overwrite b (array, cap, len), a copy.
b = []byte(result)
The copy is is discarded when you return.
What you want to do is change elements of b's array.
For example,
package main
import (
"io"
"os"
"strings"
)
func rot13(b byte) byte {
switch {
case b >= 'A' && b <= 'Z':
return 'A' + (b-'A'+13)%26
case b >= 'a' && b <= 'z':
return 'a' + (b-'a'+13)%26
}
return b
}
type rot13Reader struct {
r io.Reader
}
func (reader *rot13Reader) Read(b []byte) (int, error) {
n, err := reader.r.Read(b)
b = b[:n]
for i := range b {
b[i] = rot13(b[i])
}
return n, err
}
func main() {
s := strings.NewReader("Lbh penpxrq gur pbqr!")
r := rot13Reader{s}
io.Copy(os.Stdout, &r)
}
Playground: https://play.golang.org/p/0LDYmzrrgty
Output:
You cracked the code!
The Go Blog: Go Slices: usage and internals

I am not too sure, so please take the below with between a few grains to several pounds of salt.
First you should add an error check as early as possible:
n, err := reader.r.Read(b)
if err != nil && err == io.EOF {
fmt.Printf("\n%s, %d bytes read", err, n)
return n, err
}
With this added, the output is the one you would expect:
You cracked the code!
Lbh penpxrq gur pbqr!
EOF, 0 bytes read
The reason here is that a reader is supposed to return io.EOF in case there is nothing to read any more.
So why did you experience said strange behavior? A quick look in the source code of io.Copy reveals that b is allocated once and reused. But since b was not modified (no bytes were read) and you accessed it for reading from it, it still held the same values as before. I would argue that the underlying io.Reader should clear b in case nothing is read, as per principle of least surprise, though.

Related

Operate on a struct with lots of pointers (especially to numbers)

I have quite a few data structures that contain mostly numbers, I get the data, do a calculation and return the result.
The thing is that all of those numbers can be zero and hence, I had to switch to using pointers (*int64 or *float64) so that the default is nil and not 0.
Unfortunately, I don't know of a solution to this in Go except using pointers.
The problem comes now in the Calculate() function that is implemented for all data structures:
type X struct {
A, B, C, D, E, F *int
// and much more
Result *float64
}
func (x *X) Calculate() {
floatptr := func(f float64) *float64 { return &f }
x.Result = floatptr(float64(*x.A + *x.B + *x.C + *x.D + *x.E + *x.F))
}
This function will obviously panic if any of the data is nil. So, I wrote the functions differently that it checks for nil data before the calculation:
func (x *X) CalculateWithNilChecks() {
floatptr := func(f float64) *float64 { return &f }
if x.A == nil || x.B == nil || x.C == nil || x.D == nil || x.E == nil || x.F == nil {
return
}
x.Result = floatptr(float64(*x.A + *x.B + *x.C + *x.D + *x.E + *x.F))
}
The problem is that the data structures are quite long. Having a SUPER long if x != nil looks ugly. I was wondering if there is another (cleaner) way to doing this.
I thought of doing like in the encoding/json and just recover nil pointer dereference panics, not sure if this is cleaner TBH.
Another thought was to reflect the data structures and stop if any of the required data is nil, I don't think this should be necessary for such a simple task.
Here is a playground link for the above code
What am I missing here? Thanks!
As general solution you can unmarshal your JSON into a map of *int pointers or json.RawMessage and then use a reflection to cross check with your struct or just check it with expected number of fields.
func main() {
result := make(map[string]*int)
str := `{ "A": 1, "B": 2, "C": 3, "D": 4, "E": 5, "F": 6 }`
json.Unmarshal([]byte(str), &result)
for _, field := range result {
// Check if expect fields exists using reflection or manually
// ...
}
}
You can use reflect module and error when one of the required fields is missing.
Use this as template.
package main
import (
"fmt"
"reflect"
)
type X struct {
A, B, C, D, E, F *int
Result *float64
}
func (x *X) PrintFoo() {
fmt.Println(x.A)
}
func main() {
a := 3
x := X{A: &a}
val := reflect.ValueOf(x)
for i := 0; i < val.Type().NumField(); i++ {
field := val.Type().Field(i)
fieldType := fmt.Sprintf("%s", field.Type)
if fieldType == "*int" && val.FieldByName(field.Name).IsNil() {
fmt.Println("Missing value on field", field.Name)
}
}
}

A Tour of Go Exercise: rot13Reader, why is the reader called twice?

import (
"io"
"os"
"strings"
"fmt"
)
type rot13Reader struct {
r io.Reader
}
func (rot13 rot13Reader) Read(b []byte) (int, error){
n,err := rot13.r.Read(b)
fmt.Printf("\n%v %v\n",n, err)
const A byte ='A'
const a byte ='a'
for i,x := range b{
if x==0 {
n=i
break
}
switch {
case A<=x && x<a:
tmp := x-A
b[i] = A+((tmp+13)%26)
case a<=x && x<a+26 :
tmp := x-a
b[i] = a+((tmp+13)%26)
}
}
return n, err
}
func main() {
s := strings.NewReader("Lbh penpxrq gur pbqr!")
r := rot13Reader{s}
io.Copy(os.Stdout, &r)
}
The code above outputs
21 <nil>
You cracked the code!
0 EOF
Lbh penpxrq gur pbqr!
I have two questions :
Why is the reader being called twice (the second call returning 0 and EOF)
Why does on the second call, b is back to the original chiper? If it reads 0 character, shouldn't it still be "You cracked the code!"?

os.Read() How is work? Golang

Why if I print bs, before calling Read(), it prints nothing, but after the call file.Read(bs), it shows the inside of test.txt file. Unless bs is only argument, how Read() can Change it?
package main
import (
"os"
"fmt"
)
func main() {
file , err := os.Open("test.txt")
if err == nil {
} else {
}
stat , _ := file.Stat()
bs := make([]byte, stat.Size())
fmt.Println(string(bs))
bsf ,err := file.Read(bs)
if err != nil{
fmt.Println(err)
fmt.Println(bsf)
}
fmt.Println(string(bs))
}
Output:
(Line1)
(Line2)hi, This is Example text in test.txt file.
Unless bs is only argument, how Read() can Change it?
It seems that you may be missing basic knowledge about programming languages in general. There are different kind of "values". There are pointers (or references) and there are the "usual values".
For example:
package main
import (
"fmt"
)
func changeIt(p *int) {
*p = 9
}
func main() {
a := 1
fmt.Println(a)
changeIt(&a)
fmt.Println(a)
}
It'll print 1 9 not 1 1. *int is not an integer, but a pointer to an integer. A pointer is a value that points (references) another value. If you have a value of type pointer you get the actual value that the pointer points to by using * (which is called dereferencing):
func main() {
a := 1
b := &a
fmt.Println(b, *b)
}
b is a pointer (of type *int) that points to a. The println will print the location of a followed by the value of a which is usually something like uhm 0x10414020 1. We can also modify the value a pointer points to by using *p = ...:
func main() {
a := 1
b := &a
*b = 9
fmt.Println(b, *b, a)
}
which will print 0x10414020 9 9.
Now, []byte is a slice... slices are like pointers. When you do
func changeIt(buf []byte) {
buf[0] = 10
}
func main() {
data := []byte{1,2,3}
changeIt(data)
fmt.Println(data)
}
You're not actually passing the values [1 2 3] to changeIt but a pointer to those values. Thus here the println will show [10 2 3]. Compare this to:
func changeIt(buf [3]byte) {
buf[0] = 10
}
func main() {
data := [3]byte{1,2,3}
changeIt(data)
fmt.Println(data)
}
Which will print [1 2 3] and it will pass the values [1 2 3] and not a pointer so changeIt essentially works on a copy and the buf[0] = 10 has no effect. Remember: [n]T is an array, []T is a slice. [n]T is a "raw value" and []T is a "pointer value".

Reading an integer from standard input

How do I use the fmt.Scanf function in Go to get an integer input from the standard input?
If this can't be done using fmt.Scanf, what's the best way to read a single integer?
http://golang.org/pkg/fmt/#Scanf
All the included libraries in Go are well documented.
That being said, I believe
func main() {
var i int
_, err := fmt.Scanf("%d", &i)
}
does the trick
An alternative that can be a bit more concise is to just use fmt.Scan:
package main
import "fmt"
func main() {
var i int
fmt.Scan(&i)
fmt.Println("read number", i, "from stdin")
}
This uses reflection on the type of the argument to discover how the input should be parsed.
http://golang.org/pkg/fmt/#Scan
Here is my "Fast IO" method for reading positive integers. It could be improved with bitshifts and laying out memory in advance.
package main
import (
"io/ioutil"
"bufio"
"os"
"strconv"
)
func main() {
out := bufio.NewWriter(os.Stdout)
ints := getInts()
var T int64
T, ints = ints[0], ints[1:]
..
out.WriteString(strconv.Itoa(my_num) + "\n")
out.Flush()
}
}
func getInts() []int64 {
//assumes POSITIVE INTEGERS. Check v for '-' if you have negative.
var buf []byte
buf, _ = ioutil.ReadAll(os.Stdin)
var ints []int64
num := int64(0)
found := false
for _, v := range buf {
if '0' <= v && v <= '9' {
num = 10*num + int64(v - '0') //could use bitshifting here.
found = true
} else if found {
ints = append(ints, num)
found = false
num = 0
}
}
if found {
ints = append(ints, num)
found = false
num = 0
}
return ints
}
Golang fmt.Scan is simpler than Golang fmt.Scanf (which is simpler than Clang scanf)
If fmt.Scan errors i.e. if not nil, log & return
1 Read single variable:
import (
"fmt"
"log"
)
var i int
if _, err := fmt.Scan(&i); err != nil {
log.Print(" Scan for i failed, due to ", err)
return
}
fmt.Println(i)
2 Read multiple variables:
import (
"fmt"
"log"
)
var i, j, k int
if _, err := fmt.Scan(&i, &j, &k); err != nil {
log.Print(" Scan for i, j & k failed, due to ", err)
return
}
fmt.Println(i, j, k)
Best of luck
Example from: http://www.sortedinf.com/?q=golang-in-1-hour
You can use fmt.Scanf with a format specifier. The format specifier for the integer is %d. So you can use standard input like below.
func main() {
var someVar int
fmt.Scanf("%d", &someVar)
}
or else you can use fmt.Scan or fmt.Scanln as below.
func main() {
var someVar int
fmt.Scanln(&someVar)
}
You could also use bufio.NewReader to read an integer from the standard input.
The below program:
Prompts for an integer input
Creates a bufio.Reader to read from standard input
Reads input till it encounters a newline character '\n' (Note that this will only read a single integer. Space separated values will not work)
Removes the newline character
Converts string to int
package main
import (
"fmt"
"bufio"
"os"
"strconv"
"strings"
)
func getInt() error {
fmt.Println("Enter an integer")
userInput := bufio.NewReader(os.Stdin)
userVal, err := userInput.ReadString('\n')
if err != nil {
return err
}
input := strings.TrimSpace(userVal)
intVal, err := strconv.Atoi(input)
if err != nil {
return err
}
fmt.Printf("You entered: %d\n", intVal)
return nil
}
func main() {
getInt()
}
Why can't we just use a scanf? just like we use in C? it's working though.
package main
import "fmt"
func main() {
var i int
fmt.Scanf("%d", &i)
fmt.Println(i)
}

Looking for Go equivalent of scanf

I'm looking for the Go equivalent of scanf().
I tried with following code:
1 package main
2
3 import (
4 "scanner"
5 "os"
6 "fmt"
7 )
8
9 func main() {
10 var s scanner.Scanner
11 s.Init(os.Stdin)
12 s.Mode = scanner.ScanInts
13 tok := s.Scan()
14 for tok != scanner.EOF {
15 fmt.Printf("%d ", tok)
16 tok = s.Scan()
17 }
18 fmt.Println()
19 }
I run it with input from a text with a line of integers.
But it always output -3 -3 ...
And how to scan a line composed of a string and some integers?
Changing the mode whenever encounter a new data type?
The Package documentation:
Package scanner
A general-purpose scanner for UTF-8
encoded text.
But it seems that the scanner is not for general use.
Updated code:
func main() {
n := scanf()
fmt.Println(n)
fmt.Println(len(n))
}
func scanf() []int {
nums := new(vector.IntVector)
reader := bufio.NewReader(os.Stdin)
str, err := reader.ReadString('\n')
for err != os.EOF {
fields := strings.Fields(str)
for _, f := range fields {
i, _ := strconv.Atoi(f)
nums.Push(i)
}
str, err = reader.ReadString('\n')
}
r := make([]int, nums.Len())
for i := 0; i < nums.Len(); i++ {
r[i] = nums.At(i)
}
return r
}
Improved version:
package main
import (
"bufio"
"os"
"io"
"fmt"
"strings"
"strconv"
"container/vector"
)
func main() {
n := fscanf(os.Stdin)
fmt.Println(len(n), n)
}
func fscanf(in io.Reader) []int {
var nums vector.IntVector
reader := bufio.NewReader(in)
str, err := reader.ReadString('\n')
for err != os.EOF {
fields := strings.Fields(str)
for _, f := range fields {
if i, err := strconv.Atoi(f); err == nil {
nums.Push(i)
}
}
str, err = reader.ReadString('\n')
}
return nums
}
Your updated code was much easier to compile without the line numbers, but it was missing the package and import statements.
Looking at your code, I noticed a few things. Here's my revised version of your code.
package main
import (
"bufio"
"fmt"
"io"
"os"
"strconv"
"strings"
"container/vector"
)
func main() {
n := scanf(os.Stdin)
fmt.Println()
fmt.Println(len(n), n)
}
func scanf(in io.Reader) []int {
var nums vector.IntVector
rd := bufio.NewReader(os.Stdin)
str, err := rd.ReadString('\n')
for err != os.EOF {
fields := strings.Fields(str)
for _, f := range fields {
if i, err := strconv.Atoi(f); err == nil {
nums.Push(i)
}
}
str, err = rd.ReadString('\n')
}
return nums
}
I might want to use any input file for scanf(), not just Stdin; scanf() takes an io.Reader as a parameter.
You wrote: nums := new(vector.IntVector), where type IntVector []int. This allocates an integer slice reference named nums and initializes it to zero, then the new() function allocates an integer slice reference and initializes it to zero, and then assigns it to nums. I wrote: var nums vector.IntVector, which avoids the redundancy by simply allocating an integer slice reference named nums and initializing it to zero.
You didn't check the err value for strconv.Atoi(), which meant invalid input was converted to a zero value; I skip it.
To copy from the vector to a new slice and return the slice, you wrote:
r := make([]int, nums.Len())
for i := 0; i < nums.Len(); i++ {
r[i] = nums.At(i)
}
return r
First, I simply replaced that with an equivalent, the IntVector.Data() method: return nums.Data(). Then, I took advantage of the fact that type IntVector []int and avoided the allocation and copy by replacing that by: return nums.
Although it can be used for other things, the scanner package is designed to scan Go program text. Ints (-123), Chars('c'), Strings("str"), etc. are Go language token types.
package main
import (
"fmt"
"os"
"scanner"
"strconv"
)
func main() {
var s scanner.Scanner
s.Init(os.Stdin)
s.Error = func(s *scanner.Scanner, msg string) { fmt.Println("scan error", msg) }
s.Mode = scanner.ScanInts | scanner.ScanStrings | scanner.ScanRawStrings
for tok := s.Scan(); tok != scanner.EOF; tok = s.Scan() {
txt := s.TokenText()
fmt.Print("token:", tok, "text:", txt)
switch tok {
case scanner.Int:
si, err := strconv.Atoi64(txt)
if err == nil {
fmt.Print(" integer: ", si)
}
case scanner.String, scanner.RawString:
fmt.Print(" string: ", txt)
default:
if tok >= 0 {
fmt.Print(" unicode: ", "rune = ", tok)
} else {
fmt.Print(" ERROR")
}
}
fmt.Println()
}
}
This example always reads in a line at a time and returns the entire line as a string. If you want to parse out specific values from it you could.
package main
import (
"fmt"
"bufio"
"os"
"strings"
)
func main() {
value := Input("Please enter a value: ")
trimmed := strings.TrimSpace(value)
fmt.Printf("Hello %s!\n", trimmed)
}
func Input(str string) string {
print(str)
reader := bufio.NewReader(os.Stdin)
input, _ := reader.ReadString('\n')
return input
}
In a comment to one of my answers, you said:
From the Language Specification: "When
memory is allocated to store a value,
either through a declaration or make()
or new() call, and no explicit
initialization is provided, the memory
is given a default initialization".
Then what's the point of new()?
If we run:
package main
import ("fmt")
func main() {
var i int
var j *int
fmt.Println("i (a value) = ", i, "; j (a pointer) = ", j)
j = new(int)
fmt.Println("i (a value) = ", i, "; j (a pointer) = ", j, "; *j (a value) = ", *j)
}
The declaration var i int allocates memory to store an integer value and initializes the value to zero. The declaration var j *int allocates memory to store a pointer to an integer value and initializes the pointer to zero (a nil pointer); no memory is allocated to store an integer value. We see program output similar to:
i (a value) = 0 ; j (a pointer) = <nil>
The built-in function new takes a type T and returns a value of type *T. The memory is initialized to zero values. The statement j = new(int) allocates memory to store an integer value and initializes the value to zero, then it stores a pointer to this integer value in j. We see program output similar to:
i (a value) = 0 ; j (a pointer) = 0x7fcf913a90f0 ; *j (a value) = 0
The latest release of Go (2010-05-27) has added two functions to the fmt package: Scan() and Scanln(). They don't take any pattern string. like in C, but checks the type of the arguments instead.
package main
import (
"fmt"
"os"
"container/vector"
)
func main() {
numbers := new(vector.IntVector)
var number int
n, err := fmt.Scan(os.Stdin, &number)
for n == 1 && err == nil {
numbers.Push(number)
n, err = fmt.Scan(os.Stdin, &number)
}
fmt.Printf("%v\n", numbers.Data())
}

Resources